Background
The aim of this study was to validate a novel method of determining vena contracta area (VCA) and quantifying mitral regurgitation using multibeam high–pulse repetition frequency (HPRF) color Doppler.
Methods
The Doppler signal was isolated from the regurgitant jet, and VCA was found by summing the Doppler power from multiple beams within the vena contracta region, where calibration was done with a reference beam. In 27 patients, regurgitant volume was calculated as the product of VCA and the velocity-time integral of the regurgitant jet, measured by continuous-wave Doppler, and compared with regurgitant volume measured by magnetic resonance imaging (MRI).
Results
Spearman’s rank correlation and the 95% limits of agreement between regurgitant volume measured by MRI and by multibeam HPRF color Doppler were r s = 0.82 and −3.0 ± 26.2 mL, respectively.
Conclusion
For moderate to severe mitral regurgitation, there was good agreement between MRI and multibeam HPRF color Doppler. Agreement was lower in mild regurgitation.
The echocardiographic assessment of the severity of mitral regurgitation (MR) is challenging, because it requires the integration of different 2-dimensional (2D) Doppler parameters with inherent strengths and weaknesses. The proximal isovelocity surface area (PISA) method provides measures of mitral regurgitant flow rate, effective regurgitant orifice area (EROA) and regurgitant volume. However, the PISA method has limitations because the regurgitation is dynamic throughout systole and because it assumes a hemispheric flow convergence and a circular regurgitant orifice. A different approach to quantify MR is to look at the vena contracta, which is slightly smaller than the anatomic orifice and thus a measure of EROA. According to guidelines, a vena contracta width ≥ 7 mm and an EROA and regurgitant volume as measured by PISA ≥ 0.4 cm 2 and ≥ 60 mL, respectively, are regarded as specific signs of severe MR. Three-dimensional (3D) color flow imaging (CFI) can be used to quantify the vena contracta area (VCA) and PISA. Measuring the VCA by planimetry of 3D CFI has shown better correlation with angiographic grading than measuring the vena contracta width by 2D CFI. The shape of the VCA was investigated using 3D CFI and demonstrated to be noncircular in functional MR, resulting in poor estimation of EROA by measuring the vena contracta width.
In laminar blood flow such as in the vena contracta, the backscattered Doppler power is proportional to the volume of blood in the sample volume of an ultrasound beam. This can be calibrated to give an absolute measurement of volume flow, as was shown for measurements in arteries by Hottinger and Meindl in 1979. This principle can be used in the quantification of MR, as shown by Buck et al, who estimated volume flow through the vena contracta from the power-velocity integral by means of a single wide measurement beam and a narrow beam for calibration.
We have recently extended this principle to using multiple beams covering the vena contracta. Summing these beams provides a more homogeneous measurement compared with using a single beam. The proposed method has recently been validated in vitro as well as in computer simulations. In this study, the proposed method was used in vivo to quantify VCA semiautomatically, which was multiplied by the velocity-time integral (VTI), found independently using continuous-wave (CW) Doppler, to obtain the regurgitant volume. We used magnetic resonance imaging (MRI) as one of the reference methods for flow quantification of MR, which has previously been validated for this purpose.
Methods
Multibeam High–Pulse Repetition Frequency (HPRF) Color Doppler
The maximum velocity (the Nyquist limit) that can be resolved with CFI is typically 1 m/s for cardiac applications, which is much lower than the typical velocity of 4 to 6 m/s in a MR jet. We increased the Nyquist limit using a custom 3D HPRF color Doppler mode. With HPRF, several pulses are fired before the deep echoes from the first pulse have returned to the probe, such that exact range information is lost in exchange for an increased velocity span. However, this is not a problem, because the spurious sample volumes are outside the jet, and these signals are removed by a high-pass filter. Several ultrasound beams are spread out across the valve, as shown in the left part of Figure 1 , and 3D Doppler data are recorded from a wide but short region of interest (ROI) covering the vena contracta. (The ROI must be short radially to enable HPRF.) A composite measurement beam is made by summing the contribution from all the beams. The right part of Figure 1 shows these individual beams (black) to scale, together with the composite measurement beam (blue). A wide single measurement beam (green), as used by Buck et al, is included for comparison. The composite beam has a more homogeneous sensitivity and a higher signal-to-noise ratio than the single wide beam. Additional beams can be included to make the composite beam arbitrarily large without decreases in signal-to-noise ratio.
Blood flow through the vena contracta is known to be laminar, and for laminar blood flow, the power of the Doppler signal, P meas , is proportional to the blood volume within the sample volume. The factor of proportionality, the ratio of the Doppler power P ref to the beam area A ref , is determined from the beam with the most Doppler power, because this beam is most likely to be within the regurgitant jet or orifice, as demonstrated in the left part of Figure 1 . A computer model provides the cross-sectional area, A ref , as well as a constant, k , relating the power of the composite beam to the reference beam. Because the power-to-area ratio is constant, the VCA is found by solving
P meas VCA = k × P ref A ref .
Being a numerical method, it is independent of display parameters such as gain and tissue priority. The reason “color” is included in “3D HPRF color Doppler” is because the acquisition uses just a few beams in each direction, just like CFI, compared with pulsed wave Doppler, which typically uses 50 to 100 pulses to make one spectrum. A few samples are sufficient to separate the Doppler power from the high-velocity jet from the surroundings, enabling a reasonable frame rate. However, it is dependent on the transmit frequency and transducer size for lateral resolution, which ultimately limits minimum beam width and thus valid calibration measurements. Additionally, the grid of beams makes it possible to obtain cross-sectional images of the vena contracta, but this was not investigated in this study. Further details can be found in Hergum et al.
Data Acquisition and Processing
Equipment
A Vivid 7 Dimension (GE Vingmed Ultrasound AS, Horten, Norway) with a 3V matrix-array probe and an M4S 2D cardiac probe was used to acquire HPRF 3D color Doppler and standard 2D images. Custom software was used for postprocessing of the raw 3D Doppler data to calculate VCA. EchoPAC (GE Vingmed Ultrasound AS) was used to analyze the 2D images.
Subjects
The study was performed at the Department of Cardiology of St Olav’s Hospital (Trondheim, Norway). The primary exclusion criteria were aortic regurgitation, cardiac arrhythmias, and contraindications to MRI. Of 35 subjects with mild to severe MR who consented to participate, 3 were excluded because of poor multibeam HPRF color Doppler quality, 4 because of poor MRI quality caused by arrhythmias, and 1 because of a requirement for hemodialysis between investigations. Consequently, the study included 27 subjects, 11 women and 16 men, with a median age of 51 years (range, 28-81 years). Sixteen subjects had organic mitral valve disease or unidentified etiology, and 11 had functional regurgitation. Eccentric jets were present in 10 of 16 patients with organic MR and in 4 of 11 subjects with functional MR. All were in sinus rhythm, and the median ejection fraction measured by echocardiography was 50% (range, 23%-65%). The interval between the examinations was within 1 day in 22 of the subjects and within 1 to 5 days in 5 subjects, and there were no changes in medications. The median heart rate was 63 beats/min (range, 43-104 beats/min) during Doppler echocardiography and 64 beats/min (range, 48-110 beats/min) during MRI. The study was approved by the Regional Committee for Medical and Health Research Ethics, Norwegian Social Science Data Service, and conducted according to the Declaration of Helsinki.
Two-Dimensional Doppler Echocardiography
A standard 2D Doppler echocardiographic examination was performed, including an assessment of jet area by CFI compared with left atrial size and mitral and pulmonary vein flow. We measured the mitral regurgitant flow rate by PISA at the time of peak regurgitant jet velocity. The Nyquist limit was lowered, such that a flow convergence as close as possible to a hemisphere was shown in an apical view. The VTI and peak regurgitant jet velocity were measured using CW Doppler, and the PISA EROA and regurgitant volume were calculated using a hemispheric approach. We were unable to assess the PISA flow rate in 5 subjects. A highly qualified echocardiography observer, blinded to the MRI and multibeam HPRF color Doppler results, analyzed PISA and MR grade as follows: 1 = mild, 2 = mild to moderate, 3 = moderate to severe, and 4 = severe.
Multibeam HPRF Color Doppler
Multibeam HPRF color Doppler recordings were made in the apical view in two steps, as illustrated in Figure 2 . With the scanner in triplex mode (B mode, CFI, and pulsed-wave Doppler), the vena contracta was located by moving the pulsed-wave sample volume into the jet flow shown by CFI and adjusted to include the highest velocities. The scanner was then switched to the custom multibeam HPRF color Doppler mode, with the position of the 3D ROI centered around the pulsed-wave sample volume. Depending on the depth of the ROI, the transmit frequency and the pulse repetition frequency were set to obtain a Nyquist velocity of about 3 m/s, so that a jet flow of nearly 6 m/s could be resolved. Typically, the transmit frequency was 2.1 MHz and the pulse repetition frequency 20 kHz. In a typical subject, the depth of the ROI was at about 9 – 10 cm. Altering display parameters such as gain and tissue priority during acquisition to visualize the regurgitant jet optimally did not affect the raw 3D Doppler data we used for analysis. Each recording consisted of several cycles of real-time data, and the procedure was repeated to make sure that the vena contracta was within the ROI despite patient movement. The size of the ROI was 7 × 20 × 21 mm (radial × azimuth × elevation). Currently, the frame rate is limited to about 10 volumes/s because of the large number of transmitted ultrasound beams necessary for each acquired volume. The raw 3D Doppler data were analyzed semiautomatically to find the VCA using custom software. Radial smoothing was applied to reduce the variance of the estimates. Frames corresponding to the systolic jet were selected, and the estimate of the VCA was the median value of these measurements. Some frames were excluded according to the following criteria:
- •
there was no visible jet above the noise level,
- •
the ROI did not cover the jet area, or
- •
the vena contracta was not within the ROI.
